4D robust optimization in pencil beam scanning proton therapy for hepatocellular carcinoma

The treatment of moving targets is a challenging task using high conformal radiation techniques such as pencil beam scanning (PBS) proton therapy and requires adequate motion mitigation. Recent guidelines propose 4D robust optimization to mitigate motion artefacts in PBS therapy of thoracic malignancies. However, the availability of dosimetric analyses supporting this recommendation is limited and even non-existing for other tumour sites. The objective of this study was therefore to analyse the effectiveness of 4D robust optimization for hepatocellular carcinoma (HCC), representative for moving abdominal targets. These are usually less affected by uncertainties due to tissue heterogeneities than thoracic targets. 4D robustly optimized plans were compared with beam-specific margin plans for 6 HCC patients based on 4D dynamic accumulated doses (4DDD). 4DDD computations were conducted in RayStation with an experimentally validated routine including a site-specific beam time model. Contrary to expectations based on thoracic studies, 4D robust optimization did not yield a more homogeneous target coverage except for shallow targets close to the ribs. A clear advantage of 4D robust optimization is the sparing of normal tissue. The average dose to the normal liver could be reduced by up to 12%.